mirror of
https://github.com/BillyOutlast/flash-attention-prebuild-wheels-rocm.git
synced 2026-06-30 23:57:53 -04:00
docs: update docs for v0.6.3
This commit is contained in:
@@ -21,6 +21,8 @@
|
||||
- [Flash-Attention 2.5.6](#flash-attention-256)
|
||||
- [Flash-Attention 2.4.3](#flash-attention-243)
|
||||
- [Flash-Attention 1.0.9](#flash-attention-109)
|
||||
- [Linux arm64](#linux-arm64)
|
||||
- [Flash-Attention 2.6.3](#flash-attention-263)
|
||||
- [Windows x86_64](#windows-x86_64)
|
||||
- [Flash-Attention 2.8.3](#flash-attention-283)
|
||||
- [Flash-Attention 2.8.2](#flash-attention-282)
|
||||
@@ -959,6 +961,39 @@
|
||||
|
||||
</details>
|
||||
|
||||
## 🐧 Linux arm64
|
||||
|
||||
### Flash-Attention 2.6.3
|
||||
|
||||
<details>
|
||||
<summary>Packages for Flash-Attention 2.6.3</summary>
|
||||
|
||||
| Python | PyTorch | CUDA | package |
|
||||
| ------ | ------- | ---- | ------- |
|
||||
| 3.10 | 2.5 | 12.4 | [Download1(v0.6.3)](https://github.com/mjun0812/flash-attention-prebuild-wheels/releases/download/v0.6.3/flash_attn-2.6.3%2Bcu124torch2.5-cp310-cp310-linux_aarch64.whl) |
|
||||
| 3.10 | 2.5 | 12.8 | [Download1(v0.6.3)](https://github.com/mjun0812/flash-attention-prebuild-wheels/releases/download/v0.6.3/flash_attn-2.6.3%2Bcu128torch2.5-cp310-cp310-linux_aarch64.whl) |
|
||||
| 3.10 | 2.6 | 12.8 | [Download1(v0.6.3)](https://github.com/mjun0812/flash-attention-prebuild-wheels/releases/download/v0.6.3/flash_attn-2.6.3%2Bcu128torch2.6-cp310-cp310-linux_aarch64.whl) |
|
||||
| 3.10 | 2.7 | 12.8 | [Download1(v0.6.3)](https://github.com/mjun0812/flash-attention-prebuild-wheels/releases/download/v0.6.3/flash_attn-2.6.3%2Bcu128torch2.7-cp310-cp310-linux_aarch64.whl) |
|
||||
| 3.10 | 2.9 | 12.4 | [Download1(v0.6.3)](https://github.com/mjun0812/flash-attention-prebuild-wheels/releases/download/v0.6.3/flash_attn-2.6.3%2Bcu124torch2.9-cp310-cp310-linux_aarch64.whl) |
|
||||
| 3.10 | 2.9 | 12.8 | [Download1(v0.6.3)](https://github.com/mjun0812/flash-attention-prebuild-wheels/releases/download/v0.6.3/flash_attn-2.6.3%2Bcu128torch2.9-cp310-cp310-linux_aarch64.whl) |
|
||||
| 3.10 | 2.9 | 13.0 | [Download1(v0.6.3)](https://github.com/mjun0812/flash-attention-prebuild-wheels/releases/download/v0.6.3/flash_attn-2.6.3%2Bcu130torch2.9-cp310-cp310-linux_aarch64.whl) |
|
||||
| 3.11 | 2.5 | 12.4 | [Download1(v0.6.3)](https://github.com/mjun0812/flash-attention-prebuild-wheels/releases/download/v0.6.3/flash_attn-2.6.3%2Bcu124torch2.5-cp311-cp311-linux_aarch64.whl) |
|
||||
| 3.11 | 2.5 | 12.8 | [Download1(v0.6.3)](https://github.com/mjun0812/flash-attention-prebuild-wheels/releases/download/v0.6.3/flash_attn-2.6.3%2Bcu128torch2.5-cp311-cp311-linux_aarch64.whl) |
|
||||
| 3.11 | 2.6 | 12.8 | [Download1(v0.6.3)](https://github.com/mjun0812/flash-attention-prebuild-wheels/releases/download/v0.6.3/flash_attn-2.6.3%2Bcu128torch2.6-cp311-cp311-linux_aarch64.whl) |
|
||||
| 3.11 | 2.7 | 12.8 | [Download1(v0.6.3)](https://github.com/mjun0812/flash-attention-prebuild-wheels/releases/download/v0.6.3/flash_attn-2.6.3%2Bcu128torch2.7-cp311-cp311-linux_aarch64.whl) |
|
||||
| 3.11 | 2.9 | 12.4 | [Download1(v0.6.3)](https://github.com/mjun0812/flash-attention-prebuild-wheels/releases/download/v0.6.3/flash_attn-2.6.3%2Bcu124torch2.9-cp311-cp311-linux_aarch64.whl) |
|
||||
| 3.11 | 2.9 | 12.8 | [Download1(v0.6.3)](https://github.com/mjun0812/flash-attention-prebuild-wheels/releases/download/v0.6.3/flash_attn-2.6.3%2Bcu128torch2.9-cp311-cp311-linux_aarch64.whl) |
|
||||
| 3.11 | 2.9 | 13.0 | [Download1(v0.6.3)](https://github.com/mjun0812/flash-attention-prebuild-wheels/releases/download/v0.6.3/flash_attn-2.6.3%2Bcu130torch2.9-cp311-cp311-linux_aarch64.whl) |
|
||||
| 3.12 | 2.5 | 12.4 | [Download1(v0.6.3)](https://github.com/mjun0812/flash-attention-prebuild-wheels/releases/download/v0.6.3/flash_attn-2.6.3%2Bcu124torch2.5-cp312-cp312-linux_aarch64.whl) |
|
||||
| 3.12 | 2.5 | 12.8 | [Download1(v0.6.3)](https://github.com/mjun0812/flash-attention-prebuild-wheels/releases/download/v0.6.3/flash_attn-2.6.3%2Bcu128torch2.5-cp312-cp312-linux_aarch64.whl) |
|
||||
| 3.12 | 2.6 | 12.8 | [Download1(v0.6.3)](https://github.com/mjun0812/flash-attention-prebuild-wheels/releases/download/v0.6.3/flash_attn-2.6.3%2Bcu128torch2.6-cp312-cp312-linux_aarch64.whl) |
|
||||
| 3.12 | 2.7 | 12.8 | [Download1(v0.6.3)](https://github.com/mjun0812/flash-attention-prebuild-wheels/releases/download/v0.6.3/flash_attn-2.6.3%2Bcu128torch2.7-cp312-cp312-linux_aarch64.whl) |
|
||||
| 3.12 | 2.9 | 12.4 | [Download1(v0.6.3)](https://github.com/mjun0812/flash-attention-prebuild-wheels/releases/download/v0.6.3/flash_attn-2.6.3%2Bcu124torch2.9-cp312-cp312-linux_aarch64.whl) |
|
||||
| 3.12 | 2.9 | 12.8 | [Download1(v0.6.3)](https://github.com/mjun0812/flash-attention-prebuild-wheels/releases/download/v0.6.3/flash_attn-2.6.3%2Bcu128torch2.9-cp312-cp312-linux_aarch64.whl) |
|
||||
| 3.12 | 2.9 | 13.0 | [Download1(v0.6.3)](https://github.com/mjun0812/flash-attention-prebuild-wheels/releases/download/v0.6.3/flash_attn-2.6.3%2Bcu130torch2.9-cp312-cp312-linux_aarch64.whl) |
|
||||
|
||||
</details>
|
||||
|
||||
## 🪟 Windows x86_64
|
||||
|
||||
### Flash-Attention 2.8.3
|
||||
|
||||
@@ -1,5 +1,16 @@
|
||||
## History
|
||||
|
||||
### v0.6.3
|
||||
|
||||
[Release](https://github.com/mjun0812/flash-attention-prebuild-wheels/releases/tag/v0.6.3)
|
||||
|
||||
#### Linux arm64
|
||||
|
||||
| Flash-Attention | Python | PyTorch | CUDA |
|
||||
| --- | --- | --- | --- |
|
||||
| 2.6.3 | 3.10, 3.11, 3.12 | 2.5, 2.6, 2.7, 2.9 | 12.4, 12.8, 13.0 |
|
||||
|
||||
|
||||
### v0.5.4
|
||||
|
||||
[Release](https://github.com/mjun0812/flash-attention-prebuild-wheels/releases/tag/v0.5.4)
|
||||
|
||||
Reference in New Issue
Block a user